The central argument put forth in this proposal is that smokers' perception that they are knowledgeable about smoking-related harms plays a central role at two stages critical to communications about the harms of smoking: a) motivation to attend to and elaborate upon the information, and b) the acceptance or rejection of the information. We argue that smokers who view themselves as highly knowledgeable about the harms of smoking, which is how most smokers see themselves, have little motivation to seek out and process information about the risks of smoking. Hence, we test whether reducing smokers'perceived knowledge of smoking risks will increase attention to and processing of messages about the harms of smoking. We further propose that perceived knowledge can increase the acceptance or rejection of information via either objective (i.e., critical but unbiased) or defensive processing of information. Which process occurs is hypothesized to be moderated by smokers' perceived self-efficacy. We predict that smokers with low self-efficacy will react more defensively to communications about smoking risks than smokers with high self-efficacy, especially among those who perceive themselves as very knowledgeable. Proposed is a three-group randomized design consisting of 300 university smokers between the ages of 18-24 who smoke > 5 cigarettes/day and 300 adults smokers ages 21 and older. Smokers will complete a baseline survey assessing their perceived knowledge of the risks of smoking, self-efficacy, perceived risks and worry for smoking-related harms, and desire to quit. Two weeks after baseline, at a campus location, smokers will be randomized to take either an easy or difficult test intended to measure participants' "global knowledge" about smoking-related risks or to a . no-test control group. The two tested groups only will get veridical feedback indicating they did poorly on the test (difficult test, low perceived knowledge condition) or well (easy test, high perceived knowledge condition). All smokers will then review information on computer about risks of smoking. We will assess participants' reactions to the information (e.g., extent and favorability of cognitive responding, emotional reactions) and how these reactions affect perceptions of risk and worry for smoking-related diseases. They will be surveyed one and four months after the laboratory procedures to determine the stability of perceptions of risks, worry, knowledge, and importantly, any steps taken towards cessation (e.g., quit attempts, amount smoked, and actual cessation). [unreadable] [unreadable] [unreadable]